430 Further Notes on the Eggs of Butterfiies. 



Like these, it starts as a light yellowish-green egg, con- 

 spicuous through being so much paler than the various 

 Rutaceous leaves on which it is laid — Teclea, Toddalia, 

 Vepris, Clausena and Citrus of all species. It shortly 

 changes into a dull whitish egg with an irregular purplish 

 or brown-madder ring round its greatest circumference and 

 a dark spot in the centre. There is now a superficial 

 resemblance to the egg of Charaxes brutus, though the ring 

 of the latter eventually attains a deeper colour and the 

 shape of the egg is a little different, that of the Charaxes 

 having the usual depressed and slightly sculptured top 

 characteristic of the eggs of that genus. An ordinary 

 hand-lens makes the difference clear, and it is probable 

 that such a lens gives us much the appearance seen 

 by the small, closely-peering warbler or white-eye. In 

 P. dardanus, as in C. ethalion, I have seen a single brood 

 (out of many) in which the eggs, after developing their 

 ring, were yellow, and looked very like those of Charaxes 

 candiope. In two broods the eggs were unusually small — 

 more like those of P. demodocus. 



The resemblance of the eggs of Atella phalantha to those 

 of one or two of our commoner Acraeas is even stronger, 

 though the common appearance is again brought about by 

 a common colour and general shape with a somewhat 

 different sculpturing. Apart from the fact that Atella is 

 not so very far removed from the Acraeinae, we need to 

 know whether all Atella eggs are hke those of A. phalantha. 

 It is early days to suggest mimicry when we do not yet 

 know whether the egg-enemies that recognise by sight 

 discriminate as did the drivers. Nevertheless, the latter's 

 refusals show that a potential basis for preference certainly 

 exists, so that the possibility of mimicry is at any rate 

 worth bearing in mind ; and the results from the drivers, 

 if they should be more generally applicable, suggest that 

 if there should happen to be mimicry in any of the above 

 resemblances, mnemonic considerations may have con- 

 tributed to it in even greater degree than differences in 

 grade. I have already suggested, elsewhere (Proc. Ent. 

 Soc. 1915, p. xlii), that such considerations— the principle 

 of increased reminding-power and facilitated recognition — 

 and not Miiller's principle of the shared loss, are the real 

 basis of numerical mimicry. 



Ju-NE 2, 1916. 



